R O G E R R E V E L L E A N D H A N S E. S U E S S 24 ,lOyrs.. / 80 / /’loYrs. 20yrs 60 50 1.5% -- ___) A* UNITS of A. Fig. I. Graphic solutions of Eq. (I a) for values ~ / k ranging , from 10 to 80 years (straight lines), and of Eq. ( 5 a) for r* = 1.5 % and 1.75 % (curved lines), , 400 years (solid lines) and for average apparent C14 age of sea water ~ / k = 300 years (broken lines). Points of intersection of straight lines and curves are possible solutions of the two equations for effective reservoirs A* and S*. effect from industrial fuel combustion needs further experimental investigation. W e conclude that the exchange time z (atm) = = i l k l , defined as the time it takes on the average for a CO, molecule as a member of the atmospheric carbon reservoir to be absorbed by the sea, is of the order of magnitude of 10 years. This corresponds to a net exchange rate of the order of IO-’ mol CO, per second and square meter of the ocean surface, larger by a factor of IOO than that postulated by PLASS(1956) and smaller by a factor of IO,OOO than that deduced by DINGLE (1954) as a lower limit from numerical values of the various rate controlling constants. These are, as HUTCHINSON (1954)has forcefully pointed out, too uncertain to allow any definite conclusions. On the other hand, our exchange data give a value for the “invasion coefficient” of carbon dioxide close to that determined experimentall by BOHR (1899) for a stirred liquid sur ace. Y estimating the exchange rate of CO, between the atmosphere and the oceans: (I) that the rate constants k, and k , were not affected by a small increase of the exchangeable carbon reservoir such as that from industrial fuel combustion, and (2) that, except for that increase, no other changes in the sizes of the oceanic and atmospheric carbon reservoirs have taken place. If these assumptions were rigorously correct, the increase in atmospheric CO, due to an addition of C14 free CO, would be nearly equal to r, as given by Eq. .fs) and in table 4,and equal to the decrease in the spec& C14 activity r*, multiplied by a factor A*/A. Because of the peculiar buffer mechanism of sea water, however, the increase in the partial CO, pressure is about 10 times higher than the increase in the total CO, concentration of sea water when CO, is added and the alkalinity remains constant (BUCH,1933, see 1955)~ so that under equilibrium also HARVEY, conditions at a given alkalinity Secular variation of CO, in the atmosphere In the preceding section of this paper, two simplifylng assumptions were made when y being a numerical factor of the order of 10 Tellus IX (1957), 1